Re: Req. for views ... Win3x-95 vs LinUX

[gunnar@hafro.is (Gunnar Stefansson)]

> Received the message forwarded below from Roger, one of SEUL's  main
> hackers.   Thought  it  could  be  interesting  to read some of your
> views.
> 
> For  the  coming  year, I would like to remind that SEG participants
> _are expected_ to be active.  We  will  not move forwards if SEG's a
> passive group. Do we need a US Marine Sargent in here ? :-)

Well, having someone push people a bit around tends to help generate
discussion (sometimes).

> From: Roger R Dingledine <arma@mit.edu>
> 
> Subject: SEUL: What's wrong with Windows and right with Linux?
> 
> One of the most important  jobs  of  the  SEUL project is to provide
> convincing evidence and arguments for why each user should make  the
> change  from using Windows to using Linux.  We cannot simply produce
> better software -- we're  already  doing  that  (Redhat has links on
> their webpage to several news articles  stating  that  Hurricane  is
> easier to install and configure than NT).
> 
> There  are  a  lot  of people out there who have heard of Linux, but
> don't use it because  Windows  is  'sufficient' for them, or because
> they don't actually believe that Linux could be as  good  as  people
> claim.   We  need  to  develop a series of pages on our website that
> describe in  detail  why  Windows  is  insufficient,  and why people
> should accept Linux as *the* alternative.

Well, this is one approach, but I would like to raise a related point,
which I tend to think is more important.  I'll describe this in terms
of how I see Windows and why I use it:  This is something that I am
stuck with at the moment and I would dearly love not to have to!

The question therefore becomes:  What has to happen in order to make
it possible for people such as myself to abandon Windows permanently? 

Like almost everyone, I use Word.  This is not essential and I would
not at all mind using some other word processor such as LaTeX.
However, this is a little bit of a turn-off since Word is (at least in
principle) WYSIWIG, so it's like going back to good-old troff which I
abandoned when I finally got a word processor which could do
everything troff could and was also WYSIWYG.  Much more difficult is
the fact that I can currently read all my colleagues' Word files and I
will need to do that in the future also.  So for Linux to work it must
at least have a Word processor which can read and write Word files.
My guess is that one would point at WordPerfect from Corel, but
probably not until their new Java-based version is out of the testing
phase. 

The exact same goes for Excel.  Although this is probably not the best
spreadsheet around, there are several function/addins which are not at
present available elsewhere.  Notably this include the @Risk and
CrystalBall simulation packages and the wealth of statistical and
matrix manipulation tools.  It would be quite hard to switch over to
an environment which does not have those capabilities.  (At one point
I got sufficiently frustrated with Windows that I decided to go from
Word to LaTeX, forgetting compatibility and to go to Lotus 1-2-3, for
which @Risk is available, but no, @Risk only runs on a
Windows-compatible platform, even though the 1-2-3 program will run
e.g. on a Sun workstation)

I also use Powerpoint extensively.  This is a non-issue since I can
switch over to e.g. Corel Draw anytime.

There are a large number of other suites of Windows-based utilities
which people use, ranging from citation-handling (e.g. ProCite) to
tools for connecting their favorite palmtop computer (e.g. the
PalmPilot) and many of these may not have an obvious Linux
counterpart. 

But Windows happens to have these utilities and these utilities happen
to be used by a pretty large chunk of the scientific world.  Of course
it is a major curse to be stuck with software which only runs on one
operating system and basically on only one hardware platform - and
both the OS and the hardware platform are inferior to things which can
be obtained for the same price.

So I suspect that if one is to get anywhere with pushing Linux, that
we need to have a reasonably well-specified migration path for the
most common utilities.  The migration path needs to include a
possibility of reading and converting old files from the current
utilities. 

To summarize, to me the question is not "why" to leave Windows, but
"how". The "how" depends only on how I can get file-compatibility with
applications which I am stuck with either because I cannot find the
same functionality elsewhere or because colleagues use it.

Cheers

Gunnar